Learned latent embeddings of pulse-wave signals captured with FlowEHealth's proprietary EPG sensing technology. Replaces hand-engineered features with representations that downstream tasks (BP estimation, rhythm classification, anomaly detection) can build on. TODO: clarify pretraining objective, dataset, and which downstream tasks the embeddings have actually been benchmarked on.
EPG-LatentFlow
activeRepresentation learning over EPG pulse-wave signals.
biosignalrepresentation-learningepgpython
⊟
visualization tbd
drop a hero image at
public/projects/epg-latentflow.png and set image: "/projects/epg-latentflow.png" on this project in content/projects.ts